While we have communicated as fast as possible on the today events through supports ticket, here are the full details of the situation :
It started on Thursday with a DDoS attack on the concerned shared server. We have worked countlessly to contain this attack without affecting too much the uptime which was almost successful.
It became stronger yesterday night with more connections and it was harder to contain but the server took the load and keep serving the pages of your websites.
We worked all night over the issue and this morning it became even worse and the server wasn't able to serve the pages as it was completly saturated ( a 1900 load average was recorded for those who are aware of what it represents VS 1.2 to 6 regularly)
We kept working on banning the IPs while we started a transfer of a few websites to spread the load over several servers until we noticed that the RAID array start to fail with a greater disk access.
We stopped all the services of the server to only use the disks to transfer of the data and we have deployed a new server to transfer all the accounts to, which is still currently on going.
It would take another 6 to 12 hours maximum for all the accounts to be up and running. And another 24 hours for all the cPanel features to be enabled (clamav, softaculous, rvsitebuilder.... )
Now, it remains easy to always blame external events and deny any liability. Of course, we aren't pushing people to DDoS attack servers but the deploying a protection was in my mind over the last months and i haven't been able to choose the perfect way to go to protect our whole network. The issues is the a perfect protection doesn't exist, the cost becomes quickly huge. Changes will be made though.
Otherwise, if you open a support ticket, we will issue a one month credit for this long downtime, no question asked. It is the least we can do.